AUTHORS
Paul Buckley, PhD, Stella Bouziana, MD, PhD2, Timothy Woo3, Hanna Mohamed4, Aamina Ahmed5, Liron Barnea Slonim6, Cynthia Bishop7, Robbert Hoogeboom8, Piers E M Patten, MB ChB PhD9
AFFILIATIONS
1 King's College London, London, United Kingdom
2 King's College London, London, United Kingdom
3 King's College London, London, United Kingdom
4 King's College London, London, United Kingdom
5 King's College London, London, United Kingdom
6 King’s College Hospital NHS Foundation Trust, London, United Kingdom
7 Guys & Saint Thomas Nhs Foundation Trust, London, United Kingdom
8 King's College London, London, United Kingdom
9 King's College London, London, United Kingdom
ABSTRACT
Chronic lymphocytic leukaemia (CLL) is a dynamic and clinically heterogeneous disease sustained by a permissive lymph node (LN) tumour microenvironment (TME). Alongside genomic complexities, TME differences may contribute to this clinical heterogeneity. Proliferative aggregates in CLL-LNs (histologically described as proliferation centres) play a major role in disease pathogenesis, yet detailed single cell spatial characterisation of their architecture, and the wider LN TME, remains limited. Nonetheless, prior work has established recurrent accessory populations including T cells, macrophages and other cells. We therefore hypothesised that the CLL TME would comprise conserved architectural features across patients, including reliable presence of T cells and macrophages among both Ki67+ and Ki67- B cells. Given the clinical heterogeneity of CLL however, we investigated the extent that TME features, with a focus on proliferative aggregates, were conserved within and between patient samples.
We thus optimised a 31-plex panel to identify tumor, immune and stromal cells. LN biopsies from 7 untreated CLL patients were stained with metal isotope-tagged primary antibodies and imaged by Hyperion imaging mass cytometry, generating 18 regions of interest (ROI). Deep learning segmentation resolved 316,008 cells.
Phenotyping identified B-cell, T-cell, macrophage, stromal and wider subpopulations. First, we examined neighborhood architecture with 8 distinct CNs identified. Four of seven patients’ tumours shared a B-cell scaffold composed of four neighborhoods. These CNs differed predominantly in B-cell phenotype, rather than accessory immune content (accessory populations in these CNs varied by <4%). Furthermore, these four patients exhibited variable configuration of these CNs: indeed, the abundance of each CN comprising this shared scaffold differed considerably across patients. The remaining three patients exhibited distinct architectures, dominated respectively by a naïve CD8+ T-cell rich CN, a CD4+ memory/checkpoint environment and a CD45RA-CD27+ B cell phenotype.
We next examined areas enriched for Ki67+ B-cells, utilising density-based spatial clustering to define 241 regions (>=4 cells/30um) of Ki67+ B-cells and assessed cellular composition within 50um of these cells. Although the global abundance of Ki67+ B-cells varied considerably between patients (0.192%-9.64% / ROI), the lineages surrounding them were consistent. B cells were present in 100% (241/24) of regions, CD4+ T cells in 99%, CD8+ T cells in 93%, detectable stroma in 86% and macrophages in 80%. B cells were markedly enriched within 50um of these cells vs. background tissue (β = 1.21, FDR < 0.001) and CD4+ T cells were also elevated (β = 0.77, FDR = 0.054). We observed that Ki67hi CD4+ and CD8+ T cells were typically closer to Ki67+B cells than their Ki67lo T cell counterparts (both p<0.001). In >70% of aggregates, we observed CD45RO+CD45RA+CD27+CD4+ and CD8+ T-cells within 15um of these Ki67 clusters. Notably, the presence of most T-cell phenotypes near Ki67+B cells typically varied between patients whereas CD45RO+CD45RA+CD27+CD4+/CD8+ T cells and naïve T-cell subsets varied region-to-region within patients. These data suggest that T-cell phenotypic composition proximal to proliferative regions is patient-specific for some subsets and locally determined for others. Finally, we observed that a spatial context capturing the interface of two neighborhoods where Ki67+CD20+CD79a+ reside was the second most common context across samples.
Our data indicate an architectural scaffold within CLL, built on B-cell-rich neighborhoods infiltrated with accessory immune cells. Notably, a subset of patients had individually distinct microenvironments. Despite marked between-patient variation in proliferative burden, the presence of lineages immediately surrounding Ki67+ B cells was consistent, while the interface between diverse B-cell aggregates and Ki67+ B cells may reflect histological proliferation centres. Specific T-cell subpopulations showed strikingly variable presence and distribution, although a common CD45RO+CD45RA+CD27+ phenotype of both CD4+ and CD8+ T cells was found positioned adjacent to Ki67+ B cells. While validation using confocal microscopy of such cells is ongoing, we hypothesise that these T cell subpopulations may reflect activated and differentiated accessory subsets and note that they are well-positioned to support CLL cell survival.
DISCLOSURES
Piers E M Patten, MB ChB PhD: AstraZeneca - Honoraria; BeOne - Honoraria; BeOne - Honoraria; Astra Zeneca - Honoraria; Adaptive Biotechnologies - Honoraria; Eli Lilly - Honoraria; Lilly - Honoraria; Abbvie - Honoraria; AbbVie - Honoraria; Janssen - Honoraria; BeOne - Research Funding; AstraZeneca - Research Funding; AstraZeneca - Consultancy (Includes expert testimony); AbbVie - Consultancy (Includes expert testimony); BeOne - Consultancy (Includes expert testimony); Eli Lilly - Consultancy (Includes expert testimony)